High-end information handling systems such as, for example, network computer systems utilizing a client/server architecture conventionally have features providing a means for automatically shutting down or resetting the system when a particular event occurs or is likely to occur. However, shutting down or resetting such a system, while running in a multi-user operating configuration without first ensuring all volatile data has been saved may cause catastrophic data loss.
Known to the art are information handling systems which allow a user, such as a system administrator or the like, to set as a user definable parameter the amount of time to allow the system to flush all volatile data into a non-volatile environment before the system shuts itself down or resets itself. However, the time parameter entered by the user is typically based on his or her "best guess" of the amount of time required to perform such an operation over an average of different operating parameters and conditions. Thus, a parameter value which is too long (one which perhaps is longer than a set of physical conditions will permit) may be selected by a user causing the system to not shut down properly. This increases the possibility damage to the system. Similarly, a parameter value which is too short (one which will not allow all volatile information to be properly stored) may be selected causing a loss of data as the information handling system shuts down.
Accordingly, it would be advantageous to provide a system for automatically determining the time required for an information handling system to reach a safe shutdown and then effecting such a shutdown thereby preventing catastrophic loss of data.